{"id":"W4323266498","doi":"10.1142/s0219622023500360","title":"An Entity Extraction and Categorization Technique on Twitter Streams","year":2023,"lang":"en","type":"article","venue":"International Journal of Information Technology & Decision Making","topic":"Web Data Mining and Analysis","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Athabasca University","funders":"","keywords":"Computer science; Word2vec; Information retrieval; Categorization; Social media; Entity linking; Task (project management); Information extraction; Process (computing); Semantic Web; Annotation; World Wide Web; Knowledge base; Data science; Artificial intelligence","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007330622,0.00008987481,0.000120776,0.002659991,0.00009307859,0.0003479393,0.0008939346,0.0001242315,0.00001036711],"category_scores_gemma":[0.0003451561,0.00007787399,0.00004623166,0.0008141692,0.00003142701,0.003589316,0.0001519676,0.0002343415,0.00005217987],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00007208445,"about_ca_system_score_gemma":0.00004002578,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000002480881,"about_ca_topic_score_gemma":0.000001086324,"domain_scores_codex":[0.998651,0.00002635508,0.000552181,0.000115096,0.0005493097,0.000106104],"domain_scores_gemma":[0.9984674,0.0001059704,0.0005971658,0.0002392363,0.0005557663,0.00003449545],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00002872496,0.00003219071,0.002226127,0.000002316556,0.0000404831,0.00002832021,0.0002472661,0.001811641,0.001808958,0.02273352,0.0008119921,0.9702284],"study_design_scores_gemma":[0.002613212,0.001413968,0.04559226,0.001253827,0.00007533406,0.002874243,0.002766366,0.5358439,0.04808407,0.3236663,0.03485551,0.0009610186],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1419222,0.000006540754,0.8562764,0.0009606251,0.0004669334,0.00004436146,0.000003325564,0.0001427577,0.0001768488],"genre_scores_gemma":[0.9419628,0.00005871498,0.05771722,0.0001752228,0.00005695791,0.000004305402,0.00001377716,0.000003432973,0.000007607158],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9692674,"threshold_uncertainty_score":0.3355187,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01200378828492195,"score_gpt":0.328438379132472,"score_spread":0.3164345908475501,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}